LATE BRONZE AGE AEGEAN SHIPS AND THE PYLOS TABLETS Vn 46 AND Vn 879*

This article presents a fuller archaeological and pinacological discussion of two Pylos texts that might deal with materials for Mycenaean ship construction. It should be read in conjunction with the articles by Palaima and van Effenterre cited at the end of note *, both of which made clear the need to call upon a specialisr in nautical archaeology to provide an expert commentary on rech­nical aspects of Bronze Age ship construction and to sketch our rhe current archaeological context for the kind of «nautical» interpreta­tion of these two tablets first suggested as a possibiliry by van Eff en terre.

The fullest discussion of the alternative interpretation of V n 46 and Vn 879, that the two texts refer to materials for the construe-

A preliminary version of Hocker's contri bu tion to this article was delivered as a lectu re at

an one-day symposium on Aegean trade in the Mycenaean palatial period held in

September, 199 1 by the Program in Aegean Scripts and Prehistory of the Department of

Classics of the University of Texas , Austin . Palaima has here commented upon the ways

in which the procedures and systems fo r recording information on Linear B tablets affect

possible interpretations of items in the texts. This paper has benefitted from discussions

with Shelley Wachsmann . Hocker would also like to thank Cerna! Pulak for unpublished

information on the Uluburun wreck and his thoughts on the size and const ruction of the

ship. Pal aim a thanks N. Hirschfeld and J. R. Steffy for first inducting him in to the

mysteries of ancient ship reconstruction . Any errors that remain are ours. We use the

following abbreviations in addition ro those which are standard for readers of Minos: Basch L. Basch , le musee imaginaire de la marine antique, Athens

1987.

Palaima

Tropis I

Troprs II

van Effenterre

T. G. Palaima, «Maritime Matters in the Linear B Tablets»,

in R. Laffineur and L. Basch, eds., Thalassa: l 'egee prehisto­

rique et la mer ( = Aegaeum 7), Liege 1991, pp . 273 -309.

H. Tzalas , ed., Tropis I. !st Internationa l Symposium on

Ship Construction in Antiquity, Proceedings , Athens 1989.

H. Tzalas , ed., Tropis II ; 2nd International Symposium on

Ship Construction in Antiquity, Proceedings , Athens 1991.

H. van Effenterre, «Un navire mycenien?•, in M. Mollat ,

ed., Societes et compagnes de commerce en Orient et dans /'Ocean Indien, Paris 1970, pp. 43-53.

298 FRED HOCKER - THOMAS G. PALAIMA

tion or repair of a major room of a Mycenaean building , is L. Baumbach, «Further Thoughts on PY Vn 46», Minos 12 , 1972 , pp. 383-397. Baumbach's chief objections to van Effenterre's nautical interpretation are based on her criticisms of his imperfect understanding of ship architecture -at a time when no one, Baum­bach and van Effenterre included, had a sound knowledge of Bronze Age ship-building techniques and most interpretations proceeded by means of later historical and even modern analogies­and on her inistence that the Mycenaean term ka-pi-ni-ja cannot be related to axcicpoc;_ It has been demonstrated elsewhere (Palaima, pp . 297-298, n . 105) that the latter argument is not valid . This paper presents more probable nautical identifications of the lexical items in Vn 46 and Vn 879, thus invalidating Baumbach's first argument.

There are other problematical points in Baumbach's analysis, for example : (1) her dismissal of van Effenterre's interpretation of ta-ra-nu-we as rower's benches on the contextual grounds that the term is associated with ka-pi-ni-ja, for which supposedly «the most likely interpretation ... is still a derivative of xarcvia = xcircvrp> [ this is simply a blunt refusal to consider a reasonable alternative theory in its own terms]; (2) her proposal that ta-to -mo and ki-wo are either synonymous [ citing a 6th century poetic literary parallel, as if this were even remotely comparable to entries on consecutive lines of a Mycenaean bureaucratic document from the 13th century B.C.] or designate a free-standing vs . an attached column [ without establishing which would be which and citing as comparanda wooden columns like those from the central megaron and engaged green stone columns like those from the Treasury of Atreus: to my knowledge, the Palace of Nestor has no engaged columns, wooden or stone]; (3) her objections to interpreting e-pi- *65-ko andpe- *65-ka with the value *65 = ju [ this runs against the current evidence]; ( 4) the suggestion that other tablets, once belonging to a fuller set, are now lost and that these would have explained how the items recorded on Vn 46 and Vn 879 would have been used in building repair [ there are many comparanda for tablets without explanatory headings and the beginning of V n 46 is fragmentary, so this entire line of reasoning is either untenable or superfluous]; (5) the as­sumption that unspecified repair to a major feature of the palace ( one having columns -there is no attempt to identify a unit with two diferent kinds of single columns!- and doors, e.g., a megaron

LATE BRONZE AGE AEGEAN SHIPS AND THE PYLOS TABLETS 299

or propylon) is a more reasonable hypothesis than the construction or repair of ships which would have been essential in the palatial trade and naval defense of the territory of Pylos throughout the late Mycenaean period . In addition, Baumbach, p. 397 n . 43, claims , «That repair work was going on in the palace at the time of its destruction , is suggested by Blegen and Rawson , [Palace of N estor I, Part 1] , pp. 256 and 423». What Blegen and Rawson discuss in the passages cited is that «the front wall of Hall 65 represents a reconstruction or remodeling of a preceding structure« (p. 256) and that «[ t ]he repair of the western angle and enclosing of Courts 42 and 47 are , however, surely structures of the final phases» (p . 423) . Neither passage implies that repair work was going on at the time when the palace was destroyed, just before which the two tablets under discussion must have been written. In both cases, the remo­deling of and alterations to the existing building plan had been completed before the destruction. Repair work was not in progress. For a full treatment of how the Palace of Nestor was altered through time, see J. C. Wright, «Changes in Form and Function of the Palace at Pylos», in T. G. Palaima and C. W. Shelmerdine eds ., Py/as Comes Alive, New York 1984, pp . 19-29 . Wright , p. 28, concludes, «Perhaps the process of transforming the palace occurred piecemeal over the life of the palace» .

Again our point is not that the idea of the need for almost constant minor, and sometimes major , repairs to a complelx build­ing like the Palace of Nestor is untenable, but that: a. Baumbach has misrepresented the passages from Blegen and Rawson; b. Baumbach has provided no evidence for repair of a major unit of the palace contemporary with tablets from the destruction phase; c. the theory of building repair is no more compelling than that of ship construction and repair which Baumbach dismisses out of hand. Tablets from Pylos refer to the drafting of rowers according to an organized system, to the allocation and absence of groups of rowers, to shipbuilders, and to a coastal defense system, while texts from Knossos may monitor ships from specific regions (Palaima, pp . 285-289, 301-309). Thus it is hardly unreasonable to think that some surviving texts might refer to the component parts or raw materials used in ship-building. Our surprise that there are not more tablets like Vn 46 and Vn 879 referring to basic structural elements or component architectural parts should be equal whether the tablets are treated as referring to building repair or ship

300 FRED HOCKER - THOMAS G. PALAIMA

construction. We should note, by way of comparison, that only one tablet from Pylos (Vn 10) lists raw materials for chariot construction, despite the fuller series (from Pylos and Knossos) referring to chariot wheels and pieces of chariots in different stages of assembly. Finally, however, we agree with Baumbach, p. 383, that «[ t Jo re-examine a tablet with as many uncertainties as PY Vn 46 needs no justification». Hence the present article.

Even without the ample direct evidence of shipbuilding and seafaring in the eastern Mediterranean of the Late Bronze Age, it is abundantly clear from the wealth of «foreign» objects found on Late Bronze Age sites that long-distance exchanges of goods and ideas were an important aspect of Mycenaean, Egyptian, Cypriot, Syro­Canaanite, and Hittite culture. The sea offered the best avenue for such exchanges, and only the Hittites lacked ready access to the Mediterranean. The Amarna and Ugaritic texts further testify to the volume and variety of goods exchanged by sea, and the Tale of Wenamun provides an entertainingly tragi-comic account of sea travel at the end of the period. The ships themselves are depicted in numerous Aegean and Egyptian representations and models (with some Cypriot and Syro-Canaanite additions) from the Early Bronze Age onward.

Despite the importance of ships to a complete understanding of Late Bronze Age trade and the large number of surviving represen­tations, we know surprisingly little about them. How big were they? How many were engaged in long-distance trade? How were they built? Who built them? Governments, private individuals, or commercial groups? To what degree were they specialized for specific routes, cargos, or functions? How seaworthy were they? The representations do not tell us, although they do give us a general impression of the basic configuration of ships in the Late Bronze Age Aegean and eastern Mediterranean.

From the representations it is possible to identify two general types of Aegean vessels 1 . The first is a long, low, oared vessel with possible roots in the Early Bronze Age. Such ships are shown on the

Lucien Basch, who has studied ship iconography in the ancient world extensively and

published a thorough study of most of the known material , organizes the Minoan repre­

sentations into eight distinct groups (Basch, pp. 94-140) , but most of the representa­tions can be assigned to one of two main groups: symmetrical hulls and asymmetrical

hulls. The asymmetrical hulls are characterized by one end higher and more angular chat

the other. Further classification is based largely on provenience and decorative detail.

LATE BRONZE AG E AEGEAN SHIPS AND THE PYLOS TABLETS 301

so-called frying pans from EC II Syros and other Aegean sites 2 ,

although interpretation of these extremely schematic depictions largely consists of arguments over which end is the bow , with current scholarship preferring a «high bow» by a narrow margin 3 . By the Late Bronze Age, the possible descendants of these vessels can be seen in representations on pottery from the Cyclades and the main­land 4. These vessels are characterized by a high , plumb stem sur-

Representations from these votive objects have been carefully studied and published

since they first were discovered by Tsountas in the 1890's. The actual function of the

objects is uncertain , but it is clear chat they were not frying pans. Approximately 200

such objects are known from the Aegean and Anato lia ; most are in terracotta, but stone

and metal examples are known. See J. E. Coleman , «'Frying Pans ' of the Early Bronze

Age Aegean », AJA 89, 1985, pp. 19 1-21 9, for a recent catalogue of 124 examples and a

discussion of their form , decoration , and possible function. Of the known examples, at

least eleven are decorated with sty li zed representations of paddled or oared ships. Line

drawings of all the representations except a fragmentary example can be fo und in

Coleman , ill . 5, p. 199. Photograp hs of all the complete representations are collected

and discussed in Basch , pp . 80-89, and figs. 159-168 .

The literature on this topic is ex tensive, with arguments fo llowing two main lines. The

first concentrates on the comparison of the frying-pan ships with lacer and less ambi­

guous Aegean ship representations, such as the nearly contemporary models from Naxos

and Mochlos and LC Ia ships depicted in the West House at Akrotiri; the second follows

a functiona l argument , based on the aero- and hydrodynamics of the apparent or

hypothesized hull forms , often on the presumption chat the depicted hulls are similar to

that of an EM II clay model from Palaikastro. The first approach largely leads to a «high

bow• interpretation, although it is difficult to find comparative evidence chat is both

unambiguous and clearly related to the frying-pan representations. The second line of

reasoning is less conclusive , as most eva luations of ancient boat performance are highly

subjective, especially when the three-dimensional shape of the hull is unknown: see Y .

Vichos, «L'excremice hauce des navires ii !'age du bronze en mer Egee : La poupe ou la

proue ' Une approche naucique au probleme de !'identification des excremices des

navires cycladiques representes sur les ' poelons' de Syros,, in Tropis II , pp . 363 -370. The

Palaikastro model , although apparently similar in shape, is of little help , as the

asymmetrical hull form is suitable for propulsion in either direction and there is no

inherent indication in which direction it was intended to travel. A recent summary of

the arguments can be found in P. F. Johnston , «Bronze Age Cycladic Ships: An Overview», Temple University Aegean Symposium 7, 1982, pp. 1-8 , with a good

bibliography. The relevance of the models is discussed in P. F. Johnston , Ship and Boat Models in Ancient Greece, Annapolis 1985 , pp . 5-34. C. Broodbank , «The Longboat

and Society in the Cyclades in the Keros-Syros Culture,, AJA 93. 1989, pp. 319-337,

manages to rise above the relatively unimportant bow-seem controversy to explore the

possible soc ial and economic implications of the use of such craft .

The Pylos vesse l, on a ceramic LH Ill C 1 / 2 pyxis from Tragana , is we ll known, although

onl y recently has it been correctl y restored: G. S. Korres , [Greek title] «Representation of

a Lace Mycenaean Ship on the Pyxis from Tragana , Pylos», in Tropis I, pp . 177 -202 . The

302 FRED HOCKER -THOMAS G. PALAIMA

mounted by a bird or bird-head ornament 5 , a raised, possibly open structure in the bow, a long, low hull, possibly with bulwarks , a high, upsweeping sternpost, and another light, raised structure in the stern . Most of these characteristics persist in Geometric and later Greece and define the basic Aegean warship of Classical times . Related vessels can be seen in the mortuary temple of Ramses III at Medinet Habu in the relief of the battle with the Sea Peoples ; the ships of the Sea Peoples differ only in their symmetrical, vertical stems and stern posts , both of which carry bird-head finials 6.

A different type of vessel is depicted in a large series of Middle Bronze Age seals but is best known from the stunning miniature fresco from the West House at Akrotiri , Th era 7. The fresco shows a procession of highly decorated ships being paddled and sailed from one coastal town to another, with smaller, simpler craft rowed in their midst . The hulls are long and «crescentic», with a long, raking stem carrying a variety of ornaments; the stern is less elongated and supports a small cabin, as well as a stern projection of uncertain purpose. In the waist of each major vessel is a party of dignitaries shaded by an awning. Each ship is equipped with a mast and square sail, but only one of the vessels is under sail ; the others are propelled by a large number of paddlers leaning far over the rail. Many of

vessel resembles a large representation on a LM Ill C larnax from Gaz i-Herakleion and

small depictions on LH Ill C stirrup jars from Asine and Skyros: see Basch , pp. 140- 147.

On the recently discovered Kynos fragments , see F. Dakoronia , «Warships on Sherds of

LH Ill C Kraters from Kynos», in Tropis II , pp. 11 7-12 2.

On the continuity of ornithological motifs in Aegean ships and ship representations, see

S. Wachsmann , Seagoing Ships and Seamanship in the Late Bronze Age Levant, diss.

Hebrew University.Jerusalem 1989, pp. 144-1 59 .

H . H . Nelson et al. , Medinet Habu I: F,ar/ier Histon·cal Records of Ramses III ( Onental

Institute Publications 8) , Chicago 1930. On the relief of the naval battle, see H. H .

Nelson , «The Naval Bartle Pictured at Medinet Habu», ) NES 2, 1943, pp . 40-45 ; L.

Casson , Ships and Seamanship in the Ancient World, Princeton 197 1, pp. 36-38; S.

Wachsmann, «The Ships of the Sea Peoples», I)NA 10 , 198 1, pp. 187-220, comments

explicitly on the similarity to contemporary Aegean shi p representations.

On the miniature frescos generally, see L. Morgan , The Mimature Wall Paintings of

Thera: A Study in Aegean Culture and Iconography, Cambridge 1988 , which super­

sedes earlier reporrs ; much of what follows can be found rhere. The ship procession has

been the subject of intense debate since the discovery of the fresco in 1972 , with many

theories, some plausible and some fancifu l, advanced on the ships, their passengers, and

the occasion depicted . Fortunately, Morgan (pp . 121-1 45) has collected and summarized

most of the relevant discussion , although her references are not by any means

exhaustive.

LATE BRONZE AGE AEG EAN SHIPS AND THE PYLOS TABLETS 303

these features, particularly the basic shape of the hull and the stern cabin, or ikrion, are seen on other Aegean ship representations, such as the seals. Because the Thera fresco is extremely detailed and makes sense mechanically, it appears to provide more information than ir actually does. The hulls appear to be fuller aft than forward, both on the evidence of apparent shape and ladings, and the details of the sailing rig are both logical and within the traditions of Mediterranean single square-sail rigs. The ceremonial nature of the procession, the important passengers, the excessive decoration, the enigmatic stern projection, and the paddlers all suggest that these ships are unusual, or at least put to unusual use, and may not represent typical Aegean ships except in general form . Certainly the paddling cannot be taken as normal for seagoing ships, especially since small craft of more common appearance are rowed in the same scene. It may be that the ships are deliberately archaized ( either by their operators or the artist) by the choice of paddling and the addition of the stern projection, which can be seen on many Aegean ship representations (thus the projection seen at one end of the frying-pan ships is cited as evidence that this is the stern).

Even though it provides by far the most informative ship repre­sentations from the Bronze Age Aegean, and rivals some of the Egyptian representations for clarity and detail, the Thera fresco answers few if any of the key questions concerning Aegean ships and shipbuilding. Little if any structural information can be de­rived from the painting, nor is there any reliable indication of size , capacity, ownership, or seaworthiness . They do confirm, in more realistic fashion, earlier interpretations of the more schematic seals.

Some help, at least on structure, is offered by models from Egypt 9 . These, along with one of the few Egyptian representations

T. Gillmer , «Theories on Ship Configuration in the Bronze Age Aegean», in Tropis I,

p. 130 . Gillmer 's articles on the Theta ships, of which the above is only the latest ,

illustrate q uite graphica ll y the dangerous temptation ro over-reconstruct ion presen ted

by the frescoes. He has provided line drawings of a hull , as we ll as performance

calculations. While Gillmer is an acco mplished naval archi tect , and many of his

suggestions on the derai ls dep icted in the painting are of considera ble merit, it is

simply not poss ible ro produce a valid reconstruc tion from a two-d imensional

represencanon.

Of particul ar interest are wooden models fro m the rombs of Amenhotep 11 (M. G. A.

Reisner, Models of Sh1j;s and Boats, Cairo 1913, p . 96, figs. 348-349, no. 4944) and

Tutankhamen (H. Career, The Tomb o/Tut-A nkh-Amun Ill , London 1933, p . 60 and

fig. 63b).

304 FRED HOC KER - THOMAS G . PALAIMA

of seagoing ships (Hatshepsut's Punt ships from her mortuary tem­ple at Deir el Bahri) 10, do indicate that at least some eastern Medi­terranean ships were built on a heavy, longitudinal timber well on its way to becoming a keel. This timber appears to project inside the hull rather than outside amidships and so contributes less to directional stability under sail than a modern external keel , but it does provide significantly more longitudinal strength and stiffness than a keelplank or central strake. Unfortunately, there is no way to know if this advance had reached the Aegean by the Late Bronze Age.

The same sort of keel or proto-keel has been found on the Ulu­burun shipwreck, which has produced the first intelligible hull remains from a seagoing ship of the Bronze Age 11 . Little survives from this ship, dated to the late fourteenth century BC, but even these fragments fill in major gaps in our knowledge. The surviving cargo indicates a minimum size of 12 metric tons deadweight capac­ity 12 and a length of perhaps 15 to 18 meters. In addition, the concreted copper ingots that form the primary cargo may preserve the approximate shape of the hull amidships. The keel is a heavy timber sided (or wide) 27 cm and moulded (or thick) perhaps the same , projecting both above and below the gar boards 13. The planks , 6 cm thick, are fastened to the keel and each other by pegged mortise-and-tenon joints, spaced approximately 21 cm apart, center to center, very similar to the joints found on large Roman merchantmen of the first centuries BC/ AD 14. The woods used are silver fir (Abies sp. ) for the keel and planks, and oak (Quercus sp.) for the tenons and pegs . Prior to this discovery, the earliest evidence for pegged mortise-and-tenon joinery in Mediter-

10

I I

12

13

14

E. Naville, The Temple of Deir el Bahn· Ill: End of the Northern Half and Southern Half of the Middle Platform, London 1898, places LXXII -LXXV .

G. F. Bass, «The Construction of a Seagoing Vessel of the Lace Bronze Age», in Tropis I , pp. 25-35 , is the best source on the hull remains, although individual campaign

reports in AJA provide most of the same informat ion.

G . F. Bass and C. Pulak, personal communication . The original estimate of eigh t to

ten tons has been revised upward with the discovery in 199 1 that the cargo of copper

ingots is much larger than previously thought.

The underside of the keel is as yet inaccessible , but Cerna! Pu lak reporrs that the keel

does extend below the outboard surface of the gar board .

Compare the joinery of the Madrague de Giens vessel, an enormous amphora carrier of

the mid-first cen tury BC: A . Tchernia , P . Pomey , A . Hesnard et al. , L 'epave romaine de la Madrague de Glens (Var) (Campagnes 1972-1975) ( = Gallia Supplement 24) ,

Paris 1978 , pp . 78-80.

LATE BRONZE AGE AEGEAN SHIPS AND THE PYLOS TABLETS 305

ranean ships was from the end of the fifth century BC 15; although widely spaced, unpegged mortise-and-tenon joints were characteristic of Egyptian riverine boatbuilding since the Old Kingdom 16. Pegged mortise-and-tenon joinery was known throughout the Mediterranean by the Roman era and remained a primary feature of seagoing ship construction in the region until Byzantine times. Although the nationality and builder of the Uluburun ship are as yet unknown, the cargo , personal effects, and find spot all strongly suggest that the ship was involved in trade with the Aegean 17 and thus that Mycenaeans might well have been acquainted with this form of construction. Homer's description of the boat built by Odysseus has been interpreted as describing both this method of construction 18

and stitched or lashed construction, possibly similar to that seen on a series of Classical wrecks from the western Mediterranean 19 .

In any case, the Uluburun remains demonstrate that as early as the fourteenth century BC, some shipbuilders in the eastern Medi­terranean were not only familiar with pegged mortise-and-tenon joinery, but relied on an already sophisticated form of it in the primary structure of their hulls . This method of construction is slow and labor-intensive in comparison to other methods of wooden ship construction in use in the West today, but as we have no knowledge of alternatives available to craftsmen of the period, it is not possible to judge how great an improvement it represented at the time . Mortise-and-tenon joinery produces a strong, rigid, although heavy , hull, but it limits hull forms to relatively easy curves .

lj

16

17

18

19

C.). Eisemann and B. S. Ridgway, The Porticello Shipwreck: A Mediterranean Merchant

Vessel of 415-385 B.C , College Station, Texas 1987, pp. 10-16.

P. Lipke, The Royal Ship of Cheops (BAR International Sen·es 225/ National Mantime

Museum , Greenwich, Archaeological Senes 9), Oxford 1984, pp. 104-12 1; D . C. Patch

and C. W. Haldane , The Pharaoh's Boat at the Carnegie, Pittsburgh 1990. pp . 27-43; C.

W. Haldane, «Boat Timbers from El-Lishe: A New Method of Ancient Egyptian Hull

Construction», Man·ner's Mirror 74, 1988, pp. 141-152.

Bass (supra n. 11), «The Construction of a Seagoing Vessel of the Late Bronze Age», pp.

26-28.

L. Casson, «Odysseus' Boat (Od V, 244-257)• , AJP 85 , 1964, pp. 61-64.

S. E. Mark , «Odyssey 5.234-53 and Homeric Ship Construction: A Reappraisal •. AJA 95 .

1991 , pp. 44 1-445 argues convincingly for a laced or sewn construction, although Hocker

believes that Mark places too much emphasis on the sequence of construction. On the

other hand , Mark's flat insistence that both the Greeks and Syro-Canaanites of the Bronze

Age were familiar with pegged mortise-and-tenon joinery cannot be directly supported by

the evidence , nor can his claim that such joinery requires a society that can support full.

time, specialized shipwrights (p. 444).

306 FRED HOCKER - THOMAS G. PALAIMA

Although the Uluburun remains answer some basic structural questions about Late Bronze Age shipbuilding, they do not provide much help with larger issues of economy and the organization of shipbuilding. The Linear B tablets from Pylos , Vn 46 and Vn 879, if they in fact describe ship timbers, thus present an exciting oppor­tunity to examine Mycenaean shipbuilding from both a technolog­ical and an administrative viewpoint . The question that remains is whether these tablets actually refer to ship timbers or to components for some other structure, such as a megaron.

Because the heading is missing from both documents, there is no direct indication of the nature of the contents. Contextual inter­pretation largely concerns the reading of ka-pi-ni-ja in lines 2-4 of Vn 46. This has traditionally been read as kapnia, a sort of ceramic chimney, and the remainder of the tablet interpreted as a list of timbers for a building, but van Effenterre suggested that ka-pi-ni-ja could be related to skaphos in its later meaning of «ship». He then went on to interpret the succeeding entries as different parts of a ship and to try to make sense of the numbers associated with those parts 20 . Unfortunately, he wrote before nautical archaeology had revealed how different ancient Mediterranean shipbuilding was from modern wooden shipbuilding. While his specific interpretation requires extensive revision to accommodate the knowledge gained from the excavation of shipwrecks such as the one at Uluburun , his suggestion of a nautical context is still philologically acceptable and deserves consideration in any study of Late Bronze Age shipbuilding in the Levant.

Before the interpretation of specific terms can be tackled, something must be said about the purpose of the document as a whole , as this directly affects the relevance of the numbers of items. The tablets are, like many Linear B documents, lists kept at some sort of administrative center. If Vn 46 is a list of ship timbers, there are several possibilities, several with clear parallels in other administrative documents. The most attractive possibility is that (a) Vn 46 is a «cookbook» or «shopping list» of the timbers necessary for the construction of a single ship; this would give some meaning to

the numbers of items and suggest a rational organization of the list , based possibly on order of importance or assembly. Alternatively ,

20 Van Effenterre, pp. 43-5 3. On the problems of interpretation of ka-pi-ni-ja, see

Palaima , pp . 297 -298 .

LATE BRONZE AGE AEGEAN SHIPS AND THE PYLOS TABLETS 307

(b) the list could be a record of timber given or assigned to a ship­wright for use in the construction and repair of one or more ships 21 . It is also possible that ( c) the list is a more general order for timbers of certain kinds and sizes (thus Vn 46.2 might be read «6 timbers of sufficient size to make the a-ti-ta of a ship»). A fourth possibility is that ( d) the list represents an inventory of existing timber stocks identified by the components for which they are in­tended or suitable. Records of supp lies issued to craftsmen survive from contemporary Egyptian dockyards 22, and inventories (rather than orders) seem to be the most common form of Linear B list 23.

In any case except the first, the numbers probably do not represent the actual quantity of components needed for a single ship , but may reflect general proportions , i .e . , a small number of keel timbers, larger numbers of planks, and plenty of tenons, etc.

Regardless of the actual purpose of the document , its most im­portant contribution to the study of Mycenaean ship construction would be as a clear indication of palace interest in the material aspects of shipbuilding. Whether Vn 46 records a unique event, a special project, an ongoing program of state-controlled shipbuild­ing, or a pervasive elite influence on the shipbuilding «industry» remains to be seen, but if the document records existing stocks rather than a shopping list, it suggests a continuing program of palace involvement in a craft essential to both basic transportation and the supply and exchange of imported goods in a maritime cul­ture. Other Linear B documents from Pylos (the na-u-do-mo texts) and Mycenaean Knossos record palace supervision of the personnel involved in building and operating a sizable fleet of ships, but Vn 46 and 879, if read as nautical texts, are the first evidence of palace involvement in the management of «naval stores». This should

21

22

23

See PY Vn 10 for a clear parallel for alternative (b). Saplings and a-ko -so-ne are

delivered a-mo-te-jo-na-de for use in chariot manufacture.

The most intriguing , because of its detail , is papyrus BM 10056, originally published in

S. R. K. Glanville , «Records of a Royal Dockyard of the Time ofTuthmosis III : Papyrus

British Museum 10056•, Zeitschnft fur Agyptische Sprache und Altertumskunde 66 ,

1931 , pp . 105-121 , l"-8" ; and vol. 68, 1932, pp . 7-4 1. Like van Effenrerre's work on

Vn 46 and 879, Glanville' s interpretation of the timbers listed in the papyrus predated

the discovery of ship remains and thus depends heavily on a modern approach ro ship

construction . Now that more is known of the technical deta ils of Egyptian

shipbuilding , BM 10056 should be re-examined .

For invenrories of items described by component elements and /or materials, see the PY

Ta series, PY Tn 996 , the PY Sh series, and the KN Sk and So se ries.

308 FRED HOCKER -THOMAS G. PALAIMA

come as no surprise, as similar involvement can be seen in other crafts mentioned in the palace texts (e.g., cloth manufacture; armor, chariot and weapon production) but the scale and logistics of timber management are somewhat greater than those of raw materials such as wool.

Specific interpretations of the items listed in Vn 46 and 879 are at best problematical, as the terms lend themselves to a wide range of meanings, and the numbers cannot be relied upon as corroborat­ing evidence. That said, there are some possibilities that bear men­tioning. It should be noted that there is little correspondence between the two texts; they share only one or two words, but one of them, e-to-ki-ja, is at least one of the terms with a secure etymolog­ical explanation . What follows is ·an attempt at interpretation of the individual terms, but it must be remembered that, just as with the theory linking the items to domestic architecture, our proposals begin with an ultimately unprovable hypothesis about ka-pi-ni-ja namely that it can be read as «boat», «ship», or «hull», as opposed to the equally unprovable meaning «chimney». Due to lacunae, some of the terms are uncertain 24. The texts are presented below with the most recent readings by Emmett L. Bennett, Jr. (ELB) . Variants from the text used by van Effenterre are in italics.

We have in mind a representative ship of the period (LH III B ca. 1300-1200 BC) 25 as discussed by Hocker above. We proceed by looking for a bureaucratic logic in the texts. If these texts are orders or inventories, then a priori their lists of ship pans might be spec­ified by: (a) order of installation; (b) order of acquisition; or (c) order of structural importance of the elements of the ship(s). Since there are obvious natural sections in the texts of both Vn 46 and Vn 879, it is not illogical to think that these sections might be based on some practical logic:

Vn 46: .1 / .2-.4 / .5-.6 / . 7-.8 / .9-.10 / .11- .12 and Vn 879: .1-.2 / .3-.4.

The text of V n 879 is in its structure distinct from the text of V n 46 except for lines Vn 46.9-.10. We should note that Vn 46 and Vn

24

25 The reader should consult Palaima, pp. 296-301, for a first stage of analyzing these tablets.

The exact date of the destruction of the Palace of Nestor at Pylos has now been question­

ed by M. Popham , «Pylos: Reflections on the Date of Its Destruction and on Its Iron Age

Reoccupation», OJA 10, 1991 , pp . 315-324, who raises the possibility that the major

destruction that preserved the tablet archives took place very early in LH III B.

LATE BRONZE AGE AEGEAN SHIPS AND THE PYLOS TABLETS 309

879, although not found together, do belong to the same palaeo­graphical class (Cii), which leaves open the possibility that they might be by the same scribe or by scribes belonging to a bureau dealing with the subject of ships and their materials 26 .

Vn 46 . 0 supra mutzfa .1 pi-r~d .2 ka-pi-ni-ja , a-ti-f~, 9[ (ELB) . 3 ka-pi-ni-ja , e-ru-mi-ni-ja , 4[ .4 ka-pi-ni-ja , ta-ra-nu-we 12[

. 5 • 35-ki-no-o 81 o-pi-r~ r te-re :40[

.6 e-to-ki-ja ?3[ J-ke-te-re 140

. 7 p1-n-;a-o , ta-ra-nu-we 6

.8 qe-re-tt-rl-JO 2 me-ta-se-we 10

. 9 e-p9-wo-ke , pu-to-ro 16

.10 • 3·5-ki-no-o , pu-to-ro 100

.11 ta-to-mo , a-ro-wo , e-pi- ·65-ko

.12 c;:-1:1:1-rp1_-m-Ja 2 ki-wo-qe 1

Vn 879 .1 .2 .3 .4

a-ti( ), pe- • 65-ka ko-ni-ti-ja-ja , pe- • 65-ka e-to-ki-ja , qa-ra-de-ro pa-ke-te-re , qa-ra-de-ro

8 24

10 86

Vn 46

.1 pi-!fl3 -[ There is no real clue to the meaning of this word . We should note that, if the preserved line 1 is an entry like those that follow and not somehow part of a heading -which the tablet may or may not have had- then some item must have preced­ed those which in lines 2-4 are part of the ka-pi-ni-ja .

. 2 ka-pi-ni-ja , a-ti-tfl, 9[

26

27

As with lines 3 and 4, line 2 begins with the presumed word for boat vel. sim. (a noun of the -ta class later replaced by crxa.(l)o<;) 27 . ka-pi-ni-ja is read by van Effenterre and others as a

See Palaima , p. 300, for a brief discuss ion of the scribes and locations of tablets dea ling

with materials for chariot manufac ture and repair.

An alternative possibility, especially given the alternatives for practica l reconstruction

proposed by Hocker (see, fo r examp le , line 7) , is rha1 the term ka-pi-ni-ja refers 1101

310 FRED HOCKER - THOMAS G. PALAIMA

genitive singular, thus the line would mean «6 a-ti-ta of (for?) the boat [ or hull] ». An important question here is whether the genitive «of the boat [ or hull]» is simply part of the common name for this item or if it distinguishes a-ti-ta for a boat from a similar item used for other purposes . A modern parallel for the former might be «boatswain»; for the latter , «ship's biscuit». The use of ka-pi-ni-ja at the beginning of three succeeding lines , but not for the rest of the text, is also puzzling. Was the term only necessary for the first three items (not counting pi-r(lr[), or did the scribe decide after three lines that it was clear enough that the tablet dealt with boatbuilding? 2s. Unfortunately , a-ti-ta itself has no clear etymology. However , it should be noted that it may appear in the first preserved line of Vn 879 (see below) . If the order of the list reflects some sort of hierarchy inherent in boatbuilding, it may be that these timbers are either the first assembled or the most important, structurally . It is tempting, in either instance , to read a-ti-ta as «keel» or perhaps some other backbone component (posts?) .

. 3 ka-pi-ni-ja , e-ru-mi-ni-ja , 4(

28

The second word was read by van Effenterre as «wales» or «stringers», the principal longitudinal timbers of the ship, although one etymological derivative, elumniai, refers to beams (conceptually transverse members) , as for a roof. There is no great logical impediment to the shift in meaning , and the mention of beams or thwarts in the next line suggests that e-ru­mi-ni-ja may simply refer to a heavy, important timber such as a wale. The root connotation is of something «driven along», suggesting a longitudinal orientation . ·

generically to «boat» , but specifically to the «carved out» structural component of the

boat, i.e ., the «hull» vel sim. The use of oxci(J)o~ in the sense of 'hull of a ship ' is met

with fost in Herodotus.

See the abandonment of the repeated descriptive term a-ka-na-jo in the middle of the

fo urth entry on Cn 328, a tablet with ten su bsequent entries. W orking against this

explanation of ka-pi-ni-ja on Vn 46 are the facts that: (a) entries in lines 5-12 do not

have fo rmats parallel to the format in lines 2-4 (as is rhe case on Cn 328); (b) at least

one subsequent term in the genitive (pi-ri-ja-o in line 7) fi lls the same slot as ka-pi-nl­ja; and (3) ka-pi-nl-ja does not occur in the first p reserved line of the text. The third

point wou ld not be so critical if one could estab lish that line l were part of the

heading of Vn 46.

.5

29

30

31

. 12

LATE BRONZE AGE AEGEAN SHIPS AND THE PYLOS TABLETS 311

.4 ka-pi-ni-ja , ta-ra-nu-we 12[

ta-ra-nu-we (from 8ptivo<;) has an accepted nautical meaning of «benches» for rowers , but here probably refers to beams or thwarts, the ultimate origin of rowers ' seats. The genitive ka­pi-ni-ja and the high position in the list suggest that these ta­ra-nu-we are some of the principal timbers of the hull , rather than mere thwarts . If we accept that the schematic iconogra­phical evidence is reliable , it is entirely reasonable to assume that these beams are the principal transverse members tying the sides of the hull together. As there is as yet no hard evidence for well-developed frames in the Late Bronze Age 29 , Myce­naean ships may have resembled the Middle Kingdom Dashur boats in structure. These consist of a framel ess, edge-joined shell reinforced by a series of thwarts or beams at the sheer 30.

*35-ki-no-o 81 o-pi-r~r te-re 40[ There are no etymological indications of the meaning of the former term, but the latter has the meaning «fittings on». This could refer to fastenings of some sort, such as the tenons of mortise-and-tenon construction, or it could refer to a compo­nent added to the hull after the primary structure had been defined . Van Effenterre, who worked with a reading of the second word as o-pi-te-te-re, translated it quite directly as epi­theteres and suggested a nautical meaning of «frames» 31 . The numbers of both items in line 5 suggest either that large quan­tities of them were required or that large quantities could be acquired or stockpiled , but one should not make too much of the apparent proportion of 2: 1 32. If Vo 46 is a list of ship tim­ber, then at least one of the entries should be for planking or stock suitable for planking, probably in large quantity . Based on its position high in the list, directly after the major compo­nents of the ka-pi-ni-ja (and its recurrence with a modifier in

Some of the models show transverse banding on the interior, but Hocker cautions

aga inst making too much of this. There are no frames preserved in the excavated

portions of the Uluburun hull , but this is a relatively small area. Cern a! Pulak reporrs

that excavation in 199 1 revealed the end of a heavy timber lying at hwarrships, but it

may only be more rough lumber , as has already been fo und .

C. W . Haldane , The Dashur Boats, M.A. thesis, Texas A&M Uni versity , pp. 16-19.

Yan Effenterre, p. 49 .

Palaima, p. 299.

312 FRED HOCKER -THOMAS G. PALAIMA

line 10), a case could be made that *35-ki-no-o denotes plank­ing stock, but the exact meaning of o-pi-r~;-te-re is clouded by its broad etymological meaning. Line 5 and lines 6, 8, and 12 each list two distinct items. Does the juxtaposition of items imply some relationship between them ( cf. items and terms in PY An 607 and Vn 10)?

.6 e-to-ki-ja 73[ ]-ke-te-re 140

13

The etymological derivative of the first term ( e-to-ki-ja = en­toikhia «fittings for insertion in walls» has to do with pieces fixed in a wall . If the «wall» is here construed as the side of a ship, these 23 pieces could be any of a number of timbers: planks for the sides , or tenons, which are quite literally «fixed in the wall». They could also be thole pins, or pieces of a caprail. The second word is less enigmatic, as it may be plausibly restored as pa-ke­te-re (from ntjyvuµt), some sort of fastening elements . Since a number is given, it is assumed that these are discrete units rather than a length of cord . If such is the case, these could be tenons, tenon pegs, or other fasteners, such as were used to fasten frames and planking together. The problem with the interpretation is tied closely to the nature of the document as a whole . It this is simply an inventory, it is very thorough to count the exact number of items such as fastenings. Could this be an indication that these fastenings were somehow special, e .g. , that they were made of some expensive material, such as metal? This is not at all a necessary hypothesis . Mycenaean bookkeeping is marked by meticulous attention to exact details and numbers, since the control and handling of products and materials involved personal responsibility, and a knowledge of specific quantities was needed for allocation, delivery, or distribution . Round numbers are generally reserved for production ta,gets or taxation quotas . However, the general impression left by a study of tenons from ancient ships is that they could be easily produced in relatively large numbers, so there seems to have been no need to keep a large or long-term supply on hand. On the other hand, the Ma'agan Michael ship , wrecked on the coast of Israel in the late 5th century BC , was carrying a relatively large number of unused, partially finished tenons 33 . Perhaps this entry in Vn

E. Linder and J. P . Rosloff, personal communication .

LA TE BRONZE AGE AEGEAN SHIPS AND THE PYLOS TABLETS 313

46 does not refer to the number of finished items , but the number of pieces of wood suitable for making fasteners 34. If the list is a «cookbook», then 140 tenons or pegs are not nearly enough even to build a small vessel in the style of the Ulubu­run ship. In Hocker's opinion, this entry is the best evidence chat this list is not a complete list of the components for a single ship, as van Effenterre supposed .

. 7 p1-n-1a-o , ta-ra-nu-we 6

.8

34

jj

Here again is the word for «bench» or «beam» (from 9puvoc;) but preceded by a qualifying noun in the genitive (pi-n·-ja-o = phliaon «door-poses») with a general meaning of «poses». Even without discussing the precise nuances of pi-n·-ja-o, it is clear that these beams are differentiated from chose «of the hull» (ka-pi-ni-ja) in line 4. Thus pi-n'-ja-o must refer to some other parts of the ship requiring seats or beams. Representations indicate that Mycenaean oared ships often had raised platforms in the bow and stern (apparently directly against or fixed to the stem and sternpost) on which men could stand. The represen­tations, which are admittedly extremely schematic, suggest that these platforms were open or only lightly built. If such is the case, the beams used to support these light decks should have been of smaller scantling than those lower in the hull and thus distinguishable from them . On most later ships with multiple levels of beams and/ or decks, the beams do tend to decrease in scantling higher in the hull . Can pi-n·-ja-o be construed to refer to these structures? If so, perhaps the best reading of this line is «6 beams [ thwarts, seats] of the correct size for the castles».

qe-re-ti-ri -jo 2 me-ta-se-we 10

While there is no indication of the meaning of the first term, the second denotes intermediate pieces ( cf. the me-ta element). Van Effenterre and Palaima suggested that these might be intermediate or half-frames between floor timbers 35 . This is entirely possible, if frames existed at all. Framing timbers must be curved , and the curves for floor timbers and half-frames (the

Compare a-ko-so -ne on PY Vn 10, probably meaning «pieces of wood fo r ax les»

de livered by woodcutters ro the chario t asse mblers.

Van Effente rre, p. 49; Palaima, p . 300.

314 FRED HOCKER - THOMAS G. PALAIMA

normal components of Classical and later Mediterranean ship­building) should be distinctly different , so that even in an inventory they might be distinguished .

. 9 e-p9-wo-ke , pu-ro-ro 16

There is neither etymological nor contextual evidence for the meaning of the first word , but the similarity of lines 9 and 10 to the formulaic pattern of Vn 879 suggests the possibility that pu-to-ro is a kind of wood . If so, the lack of other such designations in the rest of the text implies that pu-to-ro was an unusual material used specifically to make e-po-wo-ke and *35-ki-no-'o . Problematical is the form pu-to-ro 36. Docs2 conjec­tures that it is perhaps a genitive plural. This would make it less likely that the item was a material and would make us think of another alternative: pu-to-ro could refer to a subsid­iary structure for which the two sets of items (e-p9-wo-ke and *35-ki-no-o) were intended . pu-to-ro would thus· be parallel to

ka-pi-ni-ja in lines 2-4 and pi-n·-ja-o in line 7 (which also provides a parallel for a subsidiary structural element in the plural) . However, the variation from the established pattern of listing the subsidiary structural elements in first position in four of the preceding lines (2-4 and 7) is troublesome; and thus the parallel with the pattern of V n 879 has more to recommend it. The form itself could be a consonant stem genitive singular (parallel to pe- *65-ka interpreted as «of pine» in Vn 879.1- .2) or it and the items pe-u-ka and qa-ra-de-ro in Vn 879 could be explained as «nominatives of rubric» or «unmarked» lexical elements in a list .

. 10 *35-ki-no-o, pu-to-ro 100

36

This is the second mention of *35-ki-no-o, here distinguished from the occurrence in line 5 by the qualifying term pu-to-ro. The quantity is similar to that in line 5, where it was suggested that it might mean planking stock . If this in fact is the case, then the planks of line 10 are either for some other purpose or of some other material.

The same problems appl y to the equall y ambiguous form qa-ra-de-ro in Vn 879 .3- .4.

for which Docs2 simply conjectures «Possibly name of a mareriaP » without com­

menting on its case or number.

LATE BRONZE AGE AEGEAN SHIPS AND THE PYLOS TABLETS 315

. 11 ta-to-mo , a-ro-wo , e-pi- * 65-ko 1

This is apparently a complex description of a single element . Etymological derivatives for the first and last terms are cna8µ6<; and trrff,oyo<;, suggesting the ideas of «standing» and «being attached». Could this refer to a maststep , which in later Medi­terranean vessels was a single, heavy timber to loca te the heel of the mast?

.1 2 c;-~u-m1.-n1-p 2 ki-wo-qe 1

As in line 3, the first entry refers to items that are «driven along», here 2 as opposed to 4 in number. The second entry means «and the column», presumably describing either a mast or possibly a heavy stanchion. If the juxtaposition of these terms does indeed imply a structural relationship, the e-ru-mi­ni-ja may be longitudinal supports for the mast, the maststep, or the partners (the structure higher in the hull that provides much of the horizontal support to the mast).

The arrangement of the text as a whole also suggests possibili­ties of interpretation. The use of ka-pi-ni-ja in lines 2-4 strongly implies some sort of internal organization. It is possible that these lines refer to the primary structure of the hull , with later lines listing components of a more general nature and for secondary structures, such as the raised fore- and afterdecks seen in contem­porary representations. Line 7 would seem to support such an inter­pretation 37 . The text would conclude with the mast (and irs supporting elements) for a single ship.

Vn 879

.1 a-ti[ ], pe-*65-ka 8

37

Based on the parallel position and similar quantity in line 2 of Vo 46, the first word might be restored as a-ti-ta. pe- *65 -ka is read «of pine», indicating that the a-ti-ta and ko-ni-ti-ja-ja of line 2 were both made of pine. Pine is a common shipbuilding wood in the Mediterranean from ancient to modern times,

Lines 9 and 10 would also support this idea, if pu-to -ro were a part of the ship ,

rather than a type of material. Bur see the d iscussion above wh ich prefers the latter

alternative .

316 FRED HOCKER - THOMAS G. PALAIMA

although its use in the Bronze Age is not yet attested by ar­chaeological evidence (Egyptian ships tend to be built of cedar and a wide variety of local woods, while the Uluburun vessel is built primarily of fir) .

. 2 ko-ni-ti-ja-ja , pe- *65 -ka 24

It should be pointed out that ko-ni-ti-ja-ja (presumably a nominative plural in -mo-) are new elements not appearing in the long list on Vn 46. This would imply: (a) that Vn 46 is not a complete «shopping list» or inventory of all necessary parts for a single ship or (b) that V n 879 contains supplemental elements, or (c) that the items on Vn 879 have to do with another kind of ship .

. 3 e-to-ki-ja , qa-ra-de-ro 10

.4 pa-ke-te-re , qa-ra-de-ro 86

It may be suggested that qa-ra-de-ro is another type of wood ( etymologically undetermined) based on parallelism with the structure of lines 1 and 2 (cf. n . 35 and related discussion). It may also be significant that the two items made of this wood , e-to-ki-ja and pa-ke-te-re from lines 3 and 4, are noted to­gether on line 6 of Vn 46 (following the reasonable restoration of the second word there) in approximately similar proportions as are listed here. If either of these terms denotes tenons for mortise-and-tenon joinery, then qa-ra-de-ro should be a hard­wood, preferably oak, the dominant material for tenons outside of Egypt; the tenons in the Uluburun ship are of oak .

To sum up, while there is no incontestable internal evidence that the tablets describe shipbuilding materials, such an interpreta­tion requires neither far-fetched explanations nor the mangling of etymology and syntax. The true nature of these tablets will only become known through the disc;overy of the heading for V n 46 or of comparable texts from other sites (like Khania: a major center of trade and industry in mid-III B that has now produced significant finds of Linear B tablets 38 ). Still the possibility that these lists describe shipbuilding forces ship historians working in the period ro

38 See E. Ha ll ager et al., «New Linear B Tablets from Kh ani a» , Kadmos 31, 1992,

61-87.

LATE BRONZE AGE AEGEAN SHIPS AND THE PYLOS TABLETS 317

re-evaluate the iconographic and archaeological evidence and to approach it in a new way 39; and historians of technology should be intrigued by the possibility that an administrative system for the management of timber supplies for shipbuilding existed over three thousand years ago.

College Station TX 77843-4352 USA Texas A&M University Nautical Archaeology Program Department of Anthropology

Austin TX 78712-1181 USA University of Texas at Austin Program in Aegean Scripts and Prehistory Department of Classics WAG 123

FRED HOCKER

TuOMAS G. PALAIMA

39 Hocker notes that after staring at these words for quite some time, he has no strong

intuitive preference for the texts referring to ship architecture rather than house architec­

ture . It should be worthwhile for experts in Mycenaean palatial archirecmre ro examine

these texts anew.